Tar analysis methods for small

scale gasification systems

…from a KTH perspective…

Klas Engvall
KTH Royal Institute of Technology

Dept. of Chemical Engineering and Technology

Stockholm, Sweden
Outline

•  Background
•  Tar analysis in a small scale gasification system
-  Offline methods
o  “Conventional” tar analysis
o  Solid Phase Absorption (SPA)
o  SPME
-  Online methods
o  FID
o  PID
-  What is the conclusion?
•  Summary

2
Background

3
 Small scale gasification systems

Pretreatment
•  Drying
•  Grinding
•  Sizing

Biomass Gas cleaning

Gasifica'on Removal: CHP
Air 700-1000 °C •  Par2cles
•  Tars

High tar Low tar

(5-50 g/Nm3) Internal combus2on engine: < 50 mg/Nm3
Gas turbines (directly fired): < 5 mg/Nm3

4
 Challenge biomass gasification tars

§  Production of condensable
polyaromatic “tars” is inherent in
most biomass gasification
processes
§  Tars foul and can plug equipment
downstream of the gasifier
§  Challenging to remove from the
produced gas
§  Reduce energy efficiency of
gasification process
§  Reports of as much as 10 % of
biomass carbon ending up in
the form of tars
5 5
 Challenge biomass gasification tars
Example internal combustion engine

Throttle valve tar deposits with water scrubbing (left) ,

oil scrubbing (right)

N. Moriconia, et al., Design and preliminary operation of a gasification plant for micro-CHP 6
with internal combustion engine and SOFC. Energy Procedia 81 ( 2015 ) 298 – 308
What is tar?

The term "tar" is vague and the definition

vary.

One definition is "organic molecules with

a molecular weight higher than that of
benzene (Mw = 78 g/mol).

7
What is tar?

Light tars
Organic compounds that can be analysed with GC as well as
HPLC. (Mw 79-300 g/mol). They are volatile and semi-
volatile aromatics and phenolics.

Heavy tar
Organic compounds with so high boiling points that they can
be analysed only by HPLC, not with GC. They are mixtures
of high molecular weight non-volatile polar
compounds (Mw ≈>300 g/mol)

Total tar = sum of light and heavy tar

8
 What is tar?

Polycyclic
Oxygenated aromatic
compounds compounds

Phenolic
compounds
and olefins

9
General scheme tar analysis

A few common steps:

1.  Sampling of the tar: Generally collected from a side stream,
including more or less complicated sampling equipment to attain a
representative sample.
2.  Storage of sample: Only valid for offline methods.
3.  Pre-preparation/conditioning of the sample:
Offline methods: The collected tars are extracted to or dissolved in
an appropriate solvent for further chemical analysis.
Online measurements: Conditioning such as drying of gas removal
of particulates etc. may be required depending on the analytical
technique.
4.  Analysis of sample: Chemical analysis of pre-prepared/conditioned
tar sample. Most common analytical techniques are based on gas
chromatography (GC) or high pressure liquid chromatography
(HPLC).

10
 All these methods!
Method Development Application Measured data Offline/ Sampling and Cost
status Online analysis time •  Large number of methods for tar
Gravimetric tar
Tar protocol/
guideline
CEN/TS pre-
standard
Laboratory
use
(Class 1)
GC-FID (Class Offline
Long sampling
and analysis
Very
expensive
measurement
2-5) time
Long sampling
Petersen
column
Portable
device
Laboratory
use
Gravimetric tar
(Class 1-5)
Offline and analysis
time
Low •  Generally:
Short sampling
Individual tar
GC-FID and Not easy to Laboratory Offline/ time and

-  For research and laboratory

compounds Expensive
GC-MS transport use Online relatively short
(Class 2-5)
analysis time
Individual tar Long sampling
HPLC
Not easy to
transport
Laboratory
use
compounds
(Class 1-5)
Offline/
Online
and analysis
time
Expensive
use
Relatively short
Individual tar
Easy to use Laboratory sampling and
-  Complex – needs expertize
SPA compounds Offline Expensive
probe use long analysis
(Class 2-5)
time
Relatively short
Easy to use Individual tar
Laboratory sampling and
SPME probe. Under compounds Offline Low
development
use
(Class 2-5)
Individual tar
long analysis
time
Relatively short
-  Several are expensive
Online tar Portable Industrial
compounds Online sampling and Expensive
analyser] device use

-  Not robust enough process

(Class 2-5) analysis time
Individual tar Short sampling
Industrial Very
MBMS Transportable compounds Online and analysis
use expensive
(Class 2-5) time
Short sampling
GC/LAMS Transportable
Laboratory
use
Individual tar
compounds
(Class 2-5)
Online
time and
relatively short
Very
expensive
-  …
analysis time
Individual tar Short sampling
Under Industrial
PID compounds Online and analysis Low
development use
(Class 2-5) time
Short sampling
Raman Not Industrial Gravimetric tar Very
Online and analysis
spectroscopy transportable use (Class 1-5) expensive
time

11
 What method to use?

•  Depends on information desired!

-  Qualitative or quantitative information?
-  Information about chemical composition?
-  R&D or industrial monitoring?
•  Reliability?
•  Costs? Pretreatment
•  Drying
•  Grinding
•  Sizing

•  …
Biomass Gas cleaning
Gasifica'on Removal: CHP
Air 700-1000 °C •  Par2cles
•  Tars

High tar Low tar

(5-50 g/Nm3) Internal combus2on engine: < 50 mg/Nm3
Gas turbines (directly fired): < 5 mg/Nm3

12
 Requirements tar analysis small scale
systems

Research Industrial monitoring/analysis

No real requirements •  Reliable - high repeatability
“A matter of need to know •  Low staffing
and costs!” •  No or low need for expertise
•  Low costs
•  Fast
Pretreatment
•  Tar composition generally
•  Drying
•  Grinding not of interest
•  Sizing

•  Preferably online
Biomass Gas cleaning
Gasifica'on Removal: CHP
Air 700-1000 °C •  Par2cles
•  Tars

High tar Low tar

(5-50 g/Nm3) Internal combus2on engine: < 50 mg/Nm3
Gas turbines (directly fired): < 5 mg/Nm3

13
Tar analysis in a small scale
gasification system

14
 Offline methods
“Conventional” tar analysis

Tar protocol developed over

several projects supported by IEA
Bioenergy Task 33, US DOE and
European Commission 1998-2005
•  Significant contributions by ECN,
VTT, KTH, DTI, BTG, NREL
Adopted as CEN standard for tar
sampling

15
CSN P CEN/TS 15439 - Biomass gasification - Tar and particles in product gases - Sampling and
analysis (2006).
 Offline methods
“Conventional” tar analysis

Procedure
•  Draw specific volume of process
gas through a filter and then a
series of cold impingers to collect
the tars
•  Evaporate solvent to measure
gravimetric (total) tars
•  Analyze tars by GC-MS to
evaluate composition
Quantitative, but very laborious

16
 Offline methods
Solid Phase Adsorption (SPA)

•  SPA sampling and analysis method was developed by

KTH in the 1990’s.

•  SPA used for measurement of the concentration (mass)

of individual light aromatic hydrocarbons and phenols.

•  The SPA-method is restricted to GC-available (GA)

compounds only.

•  These compounds are, however, significant process

markers that provide good measures of reactor
performance and gas quality.
•  At T = 900°C and above the GA-compounds roughly
correspond to the total tar amount.

Claes Brage, Qizhuang Yu, Guanxing Chen, Krister Sjöström, “Use of amino phase adsorbent 17
for biomass tar sampling and separaAon”, Fuel Vol.76, No. 2, pp. 137-142, 1997.
 Offline methods
Solid Phase Adsorption (SPA)

Solid Phase Extraction (SPE)

NH2 tube
1

4 5

7
8

10
9

11

1 = to syringe or electrical pump; 2 = adapter (polypropylene); 3 = sample reservoir; 4 = sorbent tube (PP, 1.3 OD x 7.5
cm); 5 = fritted disc (20 mm polyethylene); 6 = amino-phase sorbent (40 mm, 60 Å); 7 = rubber/silicone septum; 18
8 = septum retainer (polypropene); 9 = Tee -adapter (glass); 10 = syringe needle (stainless steel); 11 = producer gas.
 Offline methods
Solid Phase Adsorption (SPA)

572 L. Devi et al. / Renewable Energy 30 (2005) 565–587

Fig. 4. Typical chromatogram found with GC/MS analysis (inlet tar concentration).

a temperature of 300 8C. The SPA samples were analysed by a SHIMADZU QP5000 GCMS
with a WCOT fused silica column. A typical chromatogram of the inlet tar is shown in Fig. 4.
Several samples were taken at the same operating condition and the average value is
presented here. Concentrations of individual tar compounds were calculated in mg mK3 0 and
then added to get the concentration of a particular tar class. The tar compounds that were
considered for a particular class are tabulated in Table 1. Concentrations of compounds with
a higher boiling point than pyrene were determined using the calibration data of pyrene. The
heaviest compound identified was benzofluoranthene. Benzene is not considered as tar.
Light tar, e.g. toluene, was identified, but could not be measured quantitatively accurately,
especially at lower temperature. So toluene was not taken into account during total tar
calculation, but presented separately. Very high molecular weight (class 1) tars and final gas
composition could not be measured due to experimental limitations.

Sampling Sample storage Sample Chemical analysis 3.3. Catalyst characterisation

Results
Two types of additives, olivine and dolomite were tested during these experiments.

preparation Calcined dolomite is a porous catalyst; its large (internal) surface area and the presence of
oxides in its matrix (CaO, MgO) make it an active catalyst with respect to tar reduction.
Olivine is anaturally occurring silicate mineral in which magnesium and ironare embedded in
the silicate tetrahedral [10]. The properties of both the additives are tabulated in Table 3. The
BET surface area has been measured by chemisorption with ASAP. Olivine is a nonporous

Elution for aromatic GC-FID - Gas Detection limit: 2.5 mg/Nm3

material as it has an extremely low surface area. Results of mercury porosimetry for calcined

T , needle, The SPE tube is dolomite are included in Table 3 as well, for olivine mercury porosimetry could not be done

SPE-NH2 tube and 100 capped in both ends and phenolic chromatograph with (for detectable tars)
ml syringe. after sampling. compounds flame ionisation
detection
Custom made Samples stored in a
reversible SPE tube. fridge/freezer
Sampling of 100 ml in
1 min.
19
 Offline methods
Challenges of SPA Method

•  Inleakage of air, especially for •  Breakthrough of light tars (BTX)

sub-atmospheric pressure
•  Desorption of light components
systems
from SPE cartridge during
•  Using temperature high enough storage
to avoid tar condensation yet low
enough not to melt septum •  Efficient elution of aromatic and
phenolic compounds
•  Plugging of needle by septum
material •  Inability to measure heaviest
tars
•  Condensation of tars in needle
of syringe •  Consistency of procedures for
sampling and analysis
•  Undesirable heating of SPE
column during sampling due to
temperature, steam
condensation

20
Offline methods
A (biased?) comparison between SPA and “Tar guideline”

Cold solvent trapping (CST) Tradi;onal SPA

( Tar guideline ) (“KTH”)

Advantages: Advantages:
-  Gives total tar, heavy and -  Uncomplicated and fast
light tar sampling
Drawbacks: -  Low cost
-  Time consuming, sampling -  High accuracy and
as well as analysis reproducibility
-  Large solvent volumes -  Sampling and analysis can
be done separately
-  Not suitable for (very) low
tar concentraAons Drawbacks:
-  Low precision -  Not suitable for heavy tars
-  B(TX) must be analysed
within a few hours 21
 Offline methods
SPA analysis at low tar concentrations and separate BTX
analysis (KTH)
Determination of light tar in low concentrations

Adsorbent for
Benzene, Toluene
and Xylene (BTX)
Pump Flow meter
Adsorbent for other
compounds
Needle

Sample inlet

“A few other variants exists, e.g. Chalmers and ECN”

Claes Brage*, Qizhuang Yu and Krister Sjöström, A New Method for the Analysis of Heavy Tar in Raw Producer 22
Gases from biomass Gasifiers, 15th European Biomass Conference & Exhibition, 7-11 May 2007, Berlin, Germany
 Offline methods
SPA analysis combined with gravimetric tar (KTH)

Heated and isolated T - SPA

connection with SPA-
septa (left) and heavy tar
sampler (right) Heavy tar sampler Sampler mounted on atmospheric
fluidised bed gasifier
23
 Offline methods
Solid Phase Micro Extraction (SPME)
SPME device Experimental set-up KTH
Hot$gas$
Gasifier Catalyst
filter

!!°C !!°C

Exhaust Exhaust

Insulation

Heating(
tape

•  Method under development

•  Extraction of analytes from a sample matrix onto a stationary phase
(non polar) - silica fibre with 50 µm polydimethylsiloxane (PDMS)
•  Desorption of the analytes in an analytical instrument (GC).
•  Developed for low tar content analysis in e.g. syngas applications 24
 Offline methods
Solid Phase Micro Extraction (SPME)

•  Tests on real gas with sampling time 10 min.

•  The results showed that SPME method is a fast and accurate for low tar
concentrations
•  Analysis at trace levels below 0.1 mg/Nm3 (e.g., syngas production) will be
possible at 60 °C for all compounds heavier than naphthalene

Brisk report, Advanced measurement methods and operational procedures in thermochemical biomass25
conversion, D 7.6 Protocols/standards for tar measurement.
 Online methods
Flame Ionization Detector (FID)

•  Prototype developed by researcher at

University of Stuttgart.
•  The instrument is using FID as a the
detector principle.
•  The instrument can determine:
-  Total hydrocarbon concentration
-  The non-condensable hydrocarbon
concentration
-  The tar concentration

26
A. Gredinger, D. Schweitzer, H. Dieter and G. Scheffknecht, A Measurement Device for Online
Monitoring of Total Tar in Gasification Systems, J. Energy Resour. Technol 138(4), 042205
 Online methods
Flame Ionization Detector (FID)

Analyzing phase Sampling phase

Measurement principle:
Difference measurement of the organically bound carbon in the
sample gas of two sample loops with equal volume.

27
 Online methods
Flame Ionization Detector (FID)

Total hydrocarbon

Non-condensable hydrocarbons

Tar concentration

General impression:
•  Easy to use
•  Provide accurate results in comparison with ”Tar protocol”
•  The choice of a suitable tar filter material for the difference measurement was
28
identified as one of the major challenges to gain realistic results.
 Online methods
Photo Ionization Detector (PID)
A molecule with an ionization
”Method under development at KTH” potential (IP) lower than the
actual energy (E = hν) of a
photon is ionized.

Energy required to remove an

electron is different for each
compound.
Typical tar compounds require
relatively little energy

Possibly a selectivity can be

achieved. 29
 Online methods
Photo Ionization Detector (PID)
Detectable Not detectable
Energy of the light
depends on the gas
inside the lamp
Xenon = 8.4 eV
Aromatic compounds
with IP < 8.4 eV can
be detected, e.g.:
•  Naphthalene
•  Acenaphthene
•  Flourene
•  Anthracene
•  Pyrene
(eV)

Figure 17. Compounds detectable by xenon lamp (8.4 eV) 30

 Online methods
Photo Ionization Detector (PID)

116
PID in a real gas stream M. Ahmadi et al. / Fuel 113 (2013) 113–121

FIRST STAGE

FLUE GAS
SECOND STAGE
COMBUSTOR
GASIFIER
AIR

FLUIDISATION AIR ASH

VESSEL

FEED HOT
SAND CATALYTIC
BED

TO PID
DETECTOR
SAND
CHAR
ASH

RISER AIR

Fig. 5. The experimental set-up used for testing of the PID analyzer at real conditions. The PID sampling point is indicated in the figure.

Schematics of the tar measurement system for

real gas tests at BTG

31
Ahmadi M, Knoef H, Van de Beld B, Liliedahl T, Engvall K (2013) Development of a PID based on-
line tar measurement method - Proof of Concept. Fuel 113: 113-121.
 Online methods
Photo Ionization Detector (PID)

PID in a real gas stream

M. Ahmadi et al. / Fuel 113 (2013) 113–121 120
Real gas stream vs pure
119 M. Ahmadi et al. / Fuel 113 (2

naphthalene stream
be a
men
”Faulty SPA” Real gas 800
sug
thes

5. C

Naphthalene P
met
cati
proc
tar.
cati
this
mat
spo
Fig. 12. The PID signal and total tar versus time in gasification tests using different k values.
Fig. 14. Correlation of the real tar and naphthalene concentration with the PID few
signal. Real tar compounds (j) and naphthalene (N). feed
•  The PID signal follows the SPA signal at almost all different tar levels
t for real gasifiers, especially, in the small scale.
continuous PID response and the average total
down to a k = 0.13. At k < 0.13, an exponential increase of the PID
signal was observed whereas the same typetar of increase was notremains
ob-
tive
compounds similar with varying total tar concentra- mod
measured with SPA. The oxygen availability, served for naphthalene and total tar until k < 0.07. The discrepancy
tion. The correlation between the PID response and the total tar
•  Both real gas and naphthalene PID signal shows a linear correlation
gasification stage is also included in the graph.
as the ratio between the available oxygen and
between the tar concentration and the PID signal can be explained
concentration
by a dramatic increase of the benzene concentration observed at observed in this case is not perfectly linear and is
tion
is k
comparing with SPA tar content
d for complete combustion of combustible com-
raphs agree well apart from the third step when
k < 0.13. Since benzene is not defined as a not
tar itgoing
in the total tar concentration. The PID usedgeneral,
through
was not includedzero in the graph, as is displayed in Fig. 15. In
in this experiment de-
the naphthalene concentration is coming much closer to
stra
ments record a decrease in tar from the previous tects benzene since it is equipped with a 9.5 eV light source. It can
tion
describe the behavior of the PID response. The reason for this32 is pou
erintuitive and the PID is not confirming this de- however be concluded that the same general behavior displayed by
e, there were indications of faulty SPA measure-
most likely that
the tar concentration was also observed in the PID data, which is a
the different tar compounds have different PID re- that
molecular compound pyrene was not detected in sponse.
very positive result for a successful validation of theSince naphthalene has the strongest response measured in
PID method. duc
 Online methods
Photo Ionization Detector (PID)

Some observations:
•  Different response curves of the
compounds will make quantification
less accurate during analysis of real
producer gas
•  Fouling of UV lamp window with time is
an issue.
•  PID prototype test system developed
to address the window fouling problem
•  Prototype tests in progress

33
 What is the conclusion?

Gas cleaning
Biomass Gasifier
Gas cleaning
for CHP
or for fuel cell,
other?

High tar Medium tar level Low tar

(10-50 g/Nm3) (5-50 mg/Nm3) Below dew point:
Offline 0.1 mg/Nm3 at 100 °C
•  SPA •  SPA •  SPME
•  Tar protocol •  Tar protocol

•  PID •  PID
No viable technique
Online •  FID •  FID

34
Summary

35
Summary

•  Simple, low-cost, yet robust means of measuring and

characterizing tars is desirable, especially for small scale
systems
•  Impinger-based method of standard tar protocol is
relatively robust but time consuming and laborious
•  SPA method much simpler and equally as good for many
situations, but does have drawbacks
•  Continued development of SPA procedure will improve
robustness and utility of the method
•  Developed FID and PID under development are both
promising candidates for future industrial online tar
monitoring

36
 Thank you!

E-mail: kengvall@kth.se

37